NATIONAL OPTICAL ASTRONOMY OBSERVATORY Preprint Series NOAO Preprint No. 884 The Age Of The Inner Halo Globular Cluster NGC 6652 Brian Chaboyer (Department of Physics and Astronomy, Dartmouth College) Ata Sarajedini (Astronomy Department,Wesleyan University, Middletown, CT) and Taft E. Armandroff (National OpticalAstronomy Observatory) To appear in: The Astronomical Journal, December 2000 Issue September 2000 Operatedforthe National ScienceFoundationbythe Associationof Universitiesfor Researchin Astronomy,Inc. The Age of the Inner Halo Globular Cluster NGC 66521 Brian Chaboyer Department o/Physics and Astronomy, Dartmouth College, 6127 Wilder Laboratory, Hanover, NH, USA 03755-3528 [email protected] Ata Sarajedini 2 Astronomy Department, Wesleyan University, Middletown, CT 06459 ataOastro.wesleyan.edu and Taft E. Armandroff National Optical Astronomy Observatory, P.O. Box 26732, Tucson, AZ 85726 armand©noao.edu ABSTRACT HST (V, I) photometry has been obtained for the inner halo globular cluster NGC 6652. The photometry reaches ,-_ 4 mag below the turn-off and includes a well populated horizontal branch (HB). This cluster is located close to the Galactic center at RGC -- 2.0kpc with a reddening of E(V-I) = 0.15 4- 0.02 and has a metallicity of [Fe/H] -0.85. Based upon '-'*--HBAvSGB , NGC 6652 is 11.7 + 1.6 Gyr old. Using "-_--HBAvSGB precise differentialages for 47 Tuc (a thickdisk globular),MI07 and NGC 1851 (both halo clusters)were obtained. NGC 6652 appears to be the same age as 47 Tuc and NGC 1851 (within+1.2 Gyr), whilethereisa slightsuggestionthat MI07 isolderthan NGC 6652 by 2.3+ 1.5 Gyr. As this is a less than 2a result, this issue needs to be investigated further before a definitive statement regarding the relative age of M107 and NGC 6652 may be made. Subject headings: color-magnitude diagrams -- Galaxy: formation -- globular clusters: general -- globular clusters: individual (NGC 6652) 'Basedon observations made with the NASA/ESA Hubble Space Telescope obtained at the Space Telescope Science Institute, which is operated by the Association of Universities for Research in Astronomy, Incorporated, under NASA contract NAS5-26555. 2Guest User, Canadian Astronomy Data Centre, which is operated by the Dominion Astrophysical Observatory for the National Research Council of Canada's Herzberg Institute of Astrophysics. 2 1. Introduction NGC 6652 is a compact globular cluster which is projected near the Galactic center at g -- 1.53 ° and b = -11.38 ° (Harris 1996). A color magnitude diagram (CMD) for this cluster was presented by Ortolani, Bica & Barbuy (1994,hereafter OBB), who determined a number of cluster parameters, including [Fe/H] _ -0.9, E(B-V) = 0.10+0.02, < VHB >= 15.85 :t=0.04 and AV T° = 3.35±0.16. TO The small value of AVHB suggested that NGC 6652 is younger than the average Galactic halo cluster. Adopting an absolute horizontal branch (HB) magnitude of MHB -- +0.7, OBB determined RGc = 2.1 kpc. Its proximity to the Galactic center, reasonably low metallicity and small reddening make NGC 6652 somewhat unique. It is an inner halo globular cluster for which one can obtain an accurate relative age with respect to other halo clusters. Based upon its metallicity, radial velocity, HB type and position in the Galaxy, Zinn (1993) classified NGC 6652 as an old halo cluster. This is at odds with OBB's suggestion (based on the turn-off magnitude) that NGC 6652 is younger than the average halo cluster. Because NGC 6652 is a compact cluster located in a crowded field, the photometry near the turn-off obtained by OBB had a great deal of scatter leading to a large error in the age determination. For this reason, we were granted HST time to obtain a deep CMD of NGC 6652, in order to clearly delineate the turn-off region and to obtain an accurate estimate of its age. The observations and data reduction procedure are presented in §2 and the generation of the isochrones used to determine the age of NGC 6652 is discussed in §3. The analysis of the CMD is presented in §4, the discussion of the age of NGC 6652 is in §5, and the summary of the results are in §6. 2. Observations and Data Reduction NGC 6652 was observed in September 1997 with WFPC2 on HST. The cluster was centered on the PC1 chip. Observations were obtained with the F555W (V) filter (3 × 23 s and 12 × 160 s) and the F814W (I) filter (3 × 20 s and 12 × 160 s). The short exposure times were chosen to ensure good photometry at the level of the HB, while the long exposure times were picked to ensure good photometry around the turn-off and to allow for a good inter-comparison between the long and short exposure frames. Figure 1 shows the averaged, long exposure F555W image of the cluster. The WFPC2 images of NGC 6652 were divided into three sets of four long exposure (160s) F555W and F814W frames and one set of three short exposure F555W (23s) and F814W (20s) frames. Each set of four long and three short exposures in a given filter was averaged using GCOMBINE in IRAF/STSDAS with cosmic ray rejection enabled and the data quality files used to flag defective pixels. This yielded a total of 8 WFPC2 images which were input into our photometric procedure. We utilized the aperture photometry routines in DAOPHOT II (Stetson 1987, 1994) adopting an aperture radius of 2.5 pixels in the Planetary Camera and 2 pixels in the Wide Field chips. Corrections from the small aperture photometry to the standard 0.5 arcsec radius were determined using 30 to 50 relatively uncrowded stars. These gave aperture corrections with typical standard deviationsbetween0.01and0.03mag.TheV andI instrumentalmagnitudeswerematchedto form colors,andcorrectedto a 0.5arcsecradiusaperture.Thetotal instrumentalmagnitudeswerethen adjustedforexposuretimeandplacedonthestandardVI systemusingthecoefficientsin Table7of Holtzmanet al. (1995).A correctionwasalsoappliedfor the well-knownchargetransferefficiency (CTE)problemusingtheformulationofWhitmore,Heyer& Casertano(1999).Magnitudesof stars in commonbetweenthe longexposureframeswereaveragedtogetherandonly starsmeasuredon all three framepairs wereretained. The final photometryfile containsstars fainter than V=17 from the longexposureframesand brighterthan V=17 from the shortexposures(4832starsin total). In orderto ensurethat thelongandshortexposuresmagnitudeswereinternallyconsistent, the magnitudeof a numberof stars(approximately10per chip) in commonbetweenthe longand shortexposureswerecomparedand theaverageoffsetdetermined.This averageoffset(0.03mag) wasthen appliedto the short exposuremagnitudes,implying that all of the magnitudesare on the longexposurescale.The photometrydata is presentedin Table1, the full versionof whichis availablein theelectronicversionof this article. Due to its compactnature, the inner regionsof NGC 6652are crowdedon the PC1 chip. The faint (longexposure)photometrywithin r < 11.5" (250 pixels) of the cluster center showed considerable scatter around the turn-off region. For this reason, the long exposure photometry in our final CMD only includes stars with r > 11.5" from the cluster center. This trimmed data-set includes 3790 stars with good photometry. A careful inspection of the photometry obtained with the different chips showed no difference [5(V-I) < 0.005] in the location of the principal features (i.e. location of the main sequence, sub-giant branch, red giant branch and the HB) in the CMD. This indicates that the aperture corrections (applied to each chip individually) are accurate, implying that the internal errors in the photometry are very small. The external errors are set by the accuracy of the Holtzman et al. (1995) transformations to the standard system. 3. The Stellar Models and Isochrones Stellar evolution tracks for masses in the range 0.50 - 1.1 M o (in 0.05 M o increments) were calculated using the Yale stellar evolution code (Guenther et al. 1992). These models incorporate the following physics: high temperature opacities from Iglesias _ Rogers (1996); low temperature opacities from Kurucz (1991); nuclear reaction rates from Bahcall & Pinsonneault (1992) and Bahcall (1989); helium diffusion coefficients from Michaud & Proffitt (1993); and an equation of state which includes the Debye-Hfickel correction (Guenther et al. 1992). Note that this equation of state yields stellar models which are in good agreement with those derived using the OPAL equation of state (Rogers 1994; Chaboyer & Kim 1995). The surface boundary conditions were calculated using a grey T - _- relation. The stellar models employ a solar calibrated mixing length (so = 1.78). The models were typically evolved from the zero age main sequence to the upper giant branch (around My _ -1.0) in 4000 time steps. In each time step, the stellar evolution -4- equationsweresolvedwith anumericalaccuracyexceeding0.01%.Themodelsdid not includeany overshootingbeyondthe formaledgeof the convectionzones. The heavyelementcompositionof the modelswaschosento reflectthe observedabundance ratios in metal-poorstars (e.g.,Lambert1989).In particular,the a-capture elements (O, Mg, Si, S, and Ca) and Ti were enhanced by +0.4dex, while Mn was made to be deficient by the same proportions. The solar abundances (before a-enhancement) were taken from Grevesse & Noels (1993). Kurucz (low temperature) and Iglesias & Rogers (high temperature) provided us with opacities for this specific mixture. NGC 6652 has [Fe/H] -- -0.96 on the Zinn & West (1984) scale and [Fe/H] -- -0.85 on the Carretta & Gratton (1997) scale. Based upon the inclination of the red giant branch, OBB found that NGC 6652 was more metal-poor than 47 Tuc, and concluded that [Fe/H] _ -0.9. In light of these abundance determinations, stellar models and isochrones were calculated for [Fe/H] = -1.20, -1.00, -0.85, -0.70. The most metal-poor isochrone was calculated to facilitate an age comparison with somewhat more metal-poor halo clusters (see §5.2). To explore the effect of the a-element enhancement, at [Fe/H] = -0.85, scaled solar abundance models and isochrones were calculated in addition to the a-element enhanced isochrones. Our calibrated solar model had an initial solar helium abundance of Y® = 0.263 and heavy element mass fraction of Zo = 0.0182.